frontend-skills

Plugin frontend-skills from marketplace skills in malinskibeniamin/skills. Commands read from the repo's manifest — today.

Skills (200)

• accessibility

  Use when React needs ARIA, keyboard, focus, form, or nested-control accessibility.

• agent-watchdog

  Audit another agent against the original request and live evidence. Use for sessions, PRs, branches, comparisons, authorized fixes, or taking over stuck work.

• aip

  Design Google AIP resource APIs: standard methods, HTTP bindings, fields, pagination, filtering, LROs, errors and batch methods. General Protobuf contracts belong to protobuf.

• animate

  Build, review, or plan purposeful web UI motion: transitions, easing, springs, choreography, staggers, view swaps, and motion audits. Use gestures for direct manipulation; not video or native apps.

• animation-performance

  Diagnose measured web animation jank, dropped frames, busy-thread motion, per-frame React work, CSS variable invalidation, or GPU layer cost.

• animation-vocabulary

  Name a described motion effect and distinguish close animation terms. Use for 'what is this called?' questions, not implementation or general design critique.

• ask-ben

  Route Ben's frontend skills.

• ask-sonner

  Set up, style, or troubleshoot Sonner React toasts: Toaster placement, promise/loading lifecycles, duplicate messages, theme, portals, and custom content.

• bazel

  Build, test, and diagnose Bazel monorepos across Go, C++, TypeScript and React. Use for BUILD/Starlark, Bzlmod, toolchains, caching, dev loops, affected CI and release artifacts.

• bend-laws

  Prove business rules with Bend laws and check shipped TypeScript or Go against the proven model. Use for formal proof, LAWS.bend or PROOF.bend, or acceptance criteria that must hold for every input.

• blast-radius

  Find non-local breakage from a change and prove the decisive safety invariant with executable evidence. Use for blast radius, what could this break, or a suspicious small diff.

• brain-dump

  Turn raw thoughts, monologues, notes, files, or links into a multi-opportunity brief before grilling.

• break-ui

  Stress-test UI data/layout: long text, Unicode, zoom, narrow containers, empty states, and large collections. Use during diagnosis, dogfood, visual review, or explicit break-it requests.

• ci-fixes

  Diagnose and fix slow or flaky CI, inefficient jobs, and low-signal test suites. Use for CI runtime, caching, sharding, affected-only testing, runner sizing, or testing-tool modernization.

• codebase-design

  Design deep modules with small interfaces. Use when choosing seams, reducing file-hop friction, deepening modules, testing APIs, or applying shared design vocabulary.

• codex-compat

  Generate Codex hooks.json and AGENTS.md parity surfaces from the Claude hook manifest.

• ai-docs

  Use when editing Effect's ai-docs/src or regenerating its LLMS.md. Not general documentation authoring.

• babysit-pr

  Monitor, watch, or babysit a PR through CI and review. Only when monitoring is requested.

• better-accessibility

  Use for WCAG-grounded keyboard, focus, names, forms, announcements, motion and zoom checks through the existing accessibility owner.

• better-colors

  Use for palettes, semantic color tokens, theme pairs, gamut and measured contrast; preserve existing tokens and accessibility requirements.

• better-interface

  Review whole screens or flows across accessibility, layout, copy, typography, color and polish using the existing visual-review owner.

• better-layout

  Use for grouping, reading order, responsive structure, localization and RTL; preserve project density and verify actual layout failures.

• better-typography

  Use for type scales, font loading, line height, wrapping, truncation and numeric alignment within the project's type system.

• better-ui

  Use for surface geometry, optical alignment, shadows, icon states and UI polish; existing animate guidance owns motion.

• better-writing

  Use for product vocabulary, translated strings, labels, errors, empty states and confirmations through the existing ux-copy owner.

• break

  Build a temporary component stress page on explicit request; compose break-ui and prototype.

• build-design

  Use when implementing a Figma file or design image with the project's existing tokens and components through development-lifecycle.

• bundle-analysis

  Use when comparing production bundles, inspecting module composition, or maintaining bundle fixtures.

• changesets

  Use when finalizing release impact in a repository that uses Changesets. Not release publication.

• ci-maintenance

  Use when authoring or reviewing GitHub Actions workflows, composite actions, permissions, events, artifacts, or concurrency.

• cli-for-agents

  Design and review agent-friendly CLIs. Use for commands, flags, help, pipelines, errors, retries, dry runs, and machine-readable output.

• codex

  Delegate to GPT-6.1 Sol through the Codex CLI. Use for clear-spec implementation, independent review, computer use, investigation, data analysis, or token-heavy mechanical work.

• commit-push-pr

  Commit, push, and open a reviewable PR, or execute an explicitly authorized merge. Use for delivery requests; --no-pr stops after push.

• connect-query

  Build typed ConnectRPC data flows with Connect Query and Protobuf v2. Use for API calls, mutations, query hooks, transports, invalidation, or generated clients.

• contribution-triage

  Use for contribution eligibility and authorized PR moderation against the target repository's policy. Not full code review or support diagnostics.

• council

  Use for an explicit council request: three read-only reviewers from different providers. Never automatic delegation.

• create-verification-skill

  Create a project-local skill that launches, drives, observes, and cleans up the real app. Use when a repository lacks a repeatable way to prove UI, CLI, API, or service behavior.

• demo

  Create a customer-facing feature recording and publish it in a draft PR.

• dependency-maintenance

  Use when adding or moving dependencies or coordinating runtime, compiler, package-manager, image, patch, and native-build compatibility.

• development-lifecycle

  Run React, TypeScript, and UI implementation from a high-level outcome through self-verification.

• diagnosing-bugs

  Diagnosis loop for hard bugs and performance regressions. Use when asked to diagnose or debug, or when a hard bug needs a reproducible feedback loop.

• docs

  Write or review factual technical documents: READMEs, design specs, research notes, and runbooks. Use for document authoring or review, not ordinary chat, UI copy, or agent instructions.

• dogfood

  Use after each material behavior slice and before handoff or shipping to exercise runnable work at its real entrypoint.

• domain-modeling

  Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.

• e2e-testing

  Use when writing or fixing Playwright E2E specs, fixtures, browser tests, or flakes.

• effect-development

  Use when implementing or reviewing Effect compositions and patterns. For installation use effect-ts; for app migration use effect-v3-to-v4.

• effect-ts
• effect-v3-to-v4
• efficient-frontier
• eli5
• excalidraw-diagram
• explain-interface
• extend-harness
• file-pr
• frontend-invariants
• frontend-starter-kit
• full-send
• full-stop
• gestures
• go
• golang-review
• golang
• grilling
• handoff
• hill-climb
• hook-audit
• implement-spec
• improve-animations
• improve-codebase-architecture
• improve
• interface-review
• jsdocs
• launchdarkly
• learn-taste
• leaving-pr-comments
• lie-detector
• maintain-verification-skill
• make-pr-easy-to-review
• matt-ask-matt
• matt-chief-of-staff
• matt-claude-handoff
• matt-code-review
• matt-codebase-design
• matt-diagnosing-bugs
• matt-domain-modeling
• matt-git-guardrails-claude-code
• matt-grill-me
• matt-grill-with-docs
• matt-grilling
• matt-handoff
• matt-implement-spec
• matt-implement
• matt-improve-codebase-architecture
• matt-loop-me
• matt-migrate-to-shoehorn
• matt-pr
• matt-prototype
• matt-research
• matt-retro
• matt-scaffold-exercises
• matt-setup-matt-pocock-skills
• matt-setup-pre-commit
• matt-setup-ts-deep-modules
• matt-software-factory
• matt-tdd
• matt-teach
• matt-to-questionnaire
• matt-to-spec
• matt-to-tickets
• matt-triage
• matt-wait-what
• matt-wayfinder
• matt-wizard
• matt-writing-beats
• matt-writing-for-agents
• matt-writing-fragments
• matt-writing-shape
• migration-guidance
• mobile-native
• modernize
• package-development
• performance-analysis
• plan-arbiter
• plow-ahead
• postgresql
• poteto-architect
• poteto-arena
• poteto-automate-me
• poteto-benchmark-checklist
• poteto-blast-radius
• poteto-bro
• poteto-correct
• poteto-create-formal-verification-skill
• poteto-create-verification-skill
• poteto-figure-it-out
• poteto-help
• poteto-how
• poteto-in-your-own-words
• poteto-interrogate
• poteto-ketchup-video
• poteto-ketchup
• poteto-maintain-formal-verification-skill
• poteto-maintain-verification-skill
• poteto-make-bot-ui
• poteto-mode
• poteto-no-comments
• poteto-principle-attack-the-premise
• poteto-principle-boundary-discipline
• poteto-principle-build-the-lever
• poteto-principle-encode-lessons-in-structure
• poteto-principle-exhaust-the-design-space
• poteto-principle-experience-first
• poteto-principle-explain-the-number
• poteto-principle-fix-root-causes
• poteto-principle-foundational-thinking
• poteto-principle-guard-the-context-window
• poteto-principle-laziness-protocol
• poteto-principle-make-operations-idempotent
• poteto-principle-migrate-callers-then-delete-legacy-apis
• poteto-principle-minimize-reader-load
• poteto-principle-model-the-domain
• poteto-principle-never-block-on-the-human
• poteto-principle-outcome-oriented-execution
• poteto-principle-prove-it-works
• poteto-principle-redesign-from-first-principles
• poteto-principle-separate-before-serializing-shared-state
• poteto-principle-sequence-verifiable-units
• poteto-principle-subtract-before-you-add
• poteto-principle-test-behavior-not-implementation
• poteto-principle-type-system-discipline
• poteto-recall
• poteto-reflect
• poteto-reproduce-and-fix-issues
• poteto-setup-benny
• poteto-setup-pstack
• poteto-sharpen
• poteto-show-me-your-work
• poteto-statechart
• poteto-swarm
• poteto-tdd
• poteto-teach
• poteto-technical-writing
• poteto-triage-issue-reports
• poteto-typescript-best-practices
• poteto-unslop
• poteto-why
• pr-shepherd
• pr
• prime
• product
• protobuf
• prototype
• pull-requests
• quantify-impact
• read-the-damn-docs
• redpanda-ai-gateway
• registry-workflow
• release
• remove-animations
• research

Subagents (13)

• adversarial-reviewer

  Constructs concrete failure scenarios for authorized reviews. Runs for cross-model duty, diff_lines > 200, introduced P0/P1 findings, changed trust boundaries, or evidenced high-impact security/recovery risk. Outputs structured JSON findings per findings-schema.md.

• code-reviewer

  Reviews code changes for spec compliance and quality. Dispatch for two-stage PR review. Outputs structured JSON findings per findings-schema.md.

• contracts-data-reviewer

  Reviews API, schema and database compatibility plus data fidelity across serialization, precision, nullability, enums, timezones and migrations. Follows real producers and consumers.

• integration-reviewer

  Reviews harness and host integrity, content propagation, CI, dependencies and vendoring. Checks real discovery and generated outputs without treating static checks as model behavior proof.

• interface-reviewer

  Reviews rendered customer-facing surfaces, accessibility, UI copy and localized layout. Requires current visual evidence; does not invoke the explicit-only interface-review skill.

• karpathy-failure-modes
• performance-reviewer

  Reviews measured UX, bundle, animation, backend and compiler performance changes. Requires comparable evidence or an explicit structural bound; does not invent hot paths or speedup claims.

• plan-design-hat

  Design-perspective plan review. UX flow, accessibility, density, copy, and consistency. Applied inline by /grilling; dispatch requires explicit delegation.

• plan-engineering-hat

  Engineering-perspective plan review. Architecture, perf, security, tests, and dependency risk. Applied inline by /grilling; dispatch requires explicit delegation.

• plan-product-hat

  Product-perspective plan review. Asks "why, for whom, success metric, scope, reversibility" before code. Applied inline by /grilling; dispatch requires explicit delegation.

• runtime-reviewer

  Reviews concurrency and resource lifecycle across requests, workers, subscriptions and workflows. Traces cancellation, stale results, retries, ownership and cleanup from concrete interleavings.

• self-reviewer

  Reviews own session changes for quality gaps, missing tests, and simplification opportunities. Dispatched in phase 4b (Refine) before external review. Outputs structured JSON findings per findings-schema.md.

• verifier

  Runs repository-native verification and host-native surface checks after implementation. Supplies observed evidence and limits, not a substitute code review or approval.

Hooks (30)

• SessionStart

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/session-env.sh"; [ -x "$f" ] && exec "$f"; exit 0

• SessionStart

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/llm-env.sh"; [ -x "$f" ] && exec "$f"; exit 0

• SessionStart

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/pr-review-auto-hook.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PreModelSwitch

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/model-switch-router.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PostModelSwitch

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/model-switch-router.sh"; [ -x "$f" ] && exec "$f"; exit 0

• Setup

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/setup-init.sh"; [ -x "$f" ] && exec "$f"; exit 0

• UserPromptSubmit

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/quota-guard.sh"; [ -x "$f" ] && exec "$f"; exit 0

• UserPromptSubmit

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/user-prompt-context.sh"; [ -x "$f" ] && exec "$f"; exit 0

• UserPromptSubmit

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/intent-detect.sh"; [ -x "$f" ] && exec "$f"; exit 0

• UserPromptSubmit

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/pr-review-auto-hook.sh"; [ -x "$f" ] && exec "$f"; exit 0

• UserPromptExpansion

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/skill-fire-log.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PostCompact

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/post-compact-context.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PreCompact

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/pre-compact.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PostToolUseFailure

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/post-tool-failure.sh"; [ -x "$f" ] && exec "$f"; exit 0

• CwdChanged

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/cwd-changed.sh"; [ -x "$f" ] && exec "$f"; exit 0

• ConfigChange

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/config-change-guard.sh"; [ -x "$f" ] && exec "$f"; exit 0

• TaskCompleted

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/task-completed-gate.sh"; [ -x "$f" ] && exec "$f"; exit 0

• StopFailure

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/stop-failure-telemetry.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PermissionDenied

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/permission-denied-retry.sh"; [ -x "$f" ] && exec "$f"; exit 0

• Notification

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/notification-alert.sh"; [ -x "$f" ] && exec "$f"; exit 0

• FileChanged

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/file-changed-schema.sh"; [ -x "$f" ] && exec "$f"; exit 0

• FileChanged

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/file-changed-config.sh"; [ -x "$f" ] && exec "$f"; exit 0

• FileChanged

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/file-changed-env.sh"; [ -x "$f" ] && exec "$f"; exit 0

• FileChanged

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/file-changed-manifest.sh"; [ -x "$f" ] && exec "$f"; exit 0

• FileChanged

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/file-changed-deps.sh"; [ -x "$f" ] && exec "$f"; exit 0

• SessionEnd

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/session-end.sh"; [ -x "$f" ] && exec "$f"; exit 0

• SessionEnd

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/pr-review-auto-hook.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PreToolUse

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/violation-nudge.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PreToolUse

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/copyright-check.sh"; [ -x "$f" ] && exec "$f"; exit 0

• PreToolUse

  f="${CLAUDE_PLUGIN_ROOT}/.claude/hooks/bend-laws-guard.sh"; [ -x "$f" ] && exec "$f"; exit 0

Similar plugins